Early health technology assessment of tongue swab for non-sputum based pulmonary tuberculosis diagnosis in Thailand

Abstract

Background: Sputum-based diagnostic methods for pulmonary tuberculosis (PTB) are challenging for patients who cannot produce sputum. Non-sputum-based approaches, such as tongue swab (TS), can address this gap. This study conducts an early Health Technology Assessment (HTA) of TS for PTB diagnosis in Thailand.

Methods: We conducted a landscape review, stakeholder consultation, early health economic modeling, and established a Target Product Profile (TPP). The landscape review included a comprehensive literature analysis to identify gaps and unmet needs in PTB diagnosis in Thailand. Stakeholder consultations gathered insights from TB experts to validate the information. An early health economic model evaluated the cost-effectiveness of two innovative strategies: tongue swab with Loop-Mediated Isothermal Amplification (LAMP) and tongue swab with real-time polymerase chain reaction (RTPCR). The TPP outlines three target levels to guide innovators in designing effective clinical studies.

Findings: The landscape review identified the clinical workflow and reimbursement process of all PTB diagnostic tests in Thailand. The gap of tuberculosis management was around diagnosis and treatment. Stakeholders indicated that PTB detection remains inefficient due to issues such as low-test accuracy, costs, delays, drug-resistance testing, and the need for specialized laboratory techniques and personnel. TS RTPCR is the best-performing strategy, outperforming other strategies for the targeted population from the modelling analysis.

Interpretation: TS may serve as a viable alternative worth further exploration and development. An ongoing collaboration between early HTA researchers and innovators has identified valuable information for innovation development.

Funding: This work was supported by Thailand Science Research and Innovation (TSRI), Thailand, Grant Number FFB670043/0401 and Wellcome Trust grant, Grant Number 223800/Z/21/Z.

Keywords: Cost-effectiveness; Early HTA; Early health technology assessment; Economic evaluation; Economic model; Tongue swab; Tuberculosis; Tuberculosis diagnosis.

Fig. 1

Structure of decision tree model for TST, TS LAMP, TS RTPCR, sputum LAMP, sputum RTPCR strategies. Abbreviation: TST: tuberculin skin test; TS LAMP: tongue swab Loop-Mediated Isothermal Amplification; TS RTPCR: tongue swab real-time polymerase chain reaction; F/U: follow up. Group 1: suspected PTB cases aged five years and older, Group 2: suspected PTB cases younger than five years. Suspected PTB cases involve the person who exhibits at least one of the following symptoms: a persistent cough lasting more than 2 weeks, hemoptysis, an unexplained cough without a specific disease, fever, weight loss, or night sweats. Diagnostic tests for Group 1 include TS LAMP, TS RTPCR, sputum LAMP, and sputum RTPCR. For Group 2, the diagnostic tests used are TS LAMP, TS RTPCR, and TST.

Fig. 2

Structure of decision tree model for AFB and sputum Xpert strategy. Abbreviation: AFB: Acid-fast bacillus smear microscopy. Group 1: suspected PTB cases aged five years and older, Group 2: suspected PTB cases younger than five years. The AFB&Xpert strategy only involves Group 1.

Fig. 3

Cost-effectiveness Acceptability Curve for a and b. Abbreviation: TST, tuberculin skin test; TS LAMP, tongue swab Loop-Mediated Isothermal Amplification; TS RTPCR, tongue swab real-time polymerase chain reaction; AFB, acid-fast bacillus smear microscopy. a shows the Cost-effectiveness Acceptability Curve for Group 1; As the CET increases, AFB&Xpert initially show the highest cost-effectiveness. Then, TS RTPCR takes the lead, at the highest CET Sputum RTPCR becomes the most likely to be cost-effective. TS LAMP and Sputum LAMP consistently have zero probability. b presents the Cost-effectiveness Acceptability Curve for Group 2. As the CET increases, the probability of TS RTPCR being cost-effective rises, while TST decreases. Meanwhile, probability of TS LAMP remains consistently at zero.

Fig. 4

Scenario 1 and 2 analyses. Comparing Incremental Cost-Effectiveness Ratio (ICER) (a.) of sputum LAMP and sputum RTPCR to AFB&Xpert, Total Cost (b.) and Total QALY (c.) of the three strategies—AFB&Xpert, sputum LAMP and sputum RTPCR. Abbreviation: LAMP, Loop-Mediated Isothermal Amplification; RTPCR, real-time polymerase chain reaction; AFB, acid-fast bacillus smear microscopy; ICER, Incremental Cost-Effectiveness Ratio. Scenario 1: a portion of the population are unable to produce sputum for sputum-based tests. Scenario 2: AFB test requires collecting three sputum samples and conducting three tests.

Fig. 5

Expected value of perfect information (EVPI) for Group 1 and Group 2. Group 1: suspected PTB cases aged five years and older, Group 2: suspected PTB cases younger than five years. The population EVPI compared the NMB of TS LAMP, TS RTPCR, sputum LAMP, sputum RTPCR, AFB&Xpert in Group 1, and NMB of TS LAMP, TS RTPCR, TST in Group 2.

Fig. 6

The Expected Value of Perfect Parameter Information (EVPPI) for the top four parameters for Group 1. Abbreviation: TS, tongue swab; Tx, treatment; LAMP, Loop-Mediated Isothermal Amplification; AE, adverse event; TS RTPCR, tongue swab real-time polymerase chain reaction; p_spec_spu_RTPCR, specificity of sputum RTPCR; p_sen_TS_RTPCR, sensitivity of TS RTPCR; p_spec_TS_RTPCR, specificity of TS RTPCR; ratio_noearlyTX_earlyTX, hazard ratio of mortality for active PTB case without early PTB treatment during initial assessment compared with early PTB treatment. Group 1: suspected PTB cases aged five years and older.